The present invention relates to a process for polymerizing a monomer or mixtures of two or more monomers such as mixtures of ethylene and one or more comonomers, to form an interpolymer product having unique physical properties, to a process for preparing such interpolymers, and to the resulting polymer products. In another aspect, the invention relates to the articles prepared from these polymers. In certain embodiments, the inventive polymers comprise two or more regions or segments (blocks) and at least one branching center, each block being characterized by a generally uniform chemical composition. These branched, “pseudo-block” copolymers and polymeric blends comprising the same are usefully employed in the preparation of solid articles such as moldings, films, sheets, and foamed objects by molding, extruding, or other processes, and are useful as components or ingredients in adhesives, laminates, polymeric blends, and other end uses. The resulting products are used in the manufacture of components for automobiles, such as profiles, bumpers and trim parts; packaging materials; electric cable insulation, and other applications. Blends of polymers according to the invention with one or more natural or synthetic polymers may be prepared as well.
It has long been known that polymers containing a block-type structure often have superior properties compared to random copolymers and blends. For example, triblock copolymers of styrene and butadiene (SBS) and hydrogenated versions of the same (SEBS) have an excellent combination of heat resistance and elasticity. Other block copolymers are also known in the art. Generally, block copolymers known as thermoplastic elastomers (TPE) have desirable properties due to the presence of “soft” or elastomeric block segments connecting “hard” either crystallizable or glassy blocks in the same polymer. At temperatures up to the melt temperature or glass transition temperature of the hard segments, the polymers demonstrate elastomeric character. At higher temperatures, the polymers become flowable, exhibiting thermoplastic behavior. Known methods of preparing block copolymers include anionic polymerization and controlled free radical polymerization. Unfortunately, these methods of preparing block copolymers require sequential monomer addition with polymerization to relative completeness and the types of monomers that can be usefully employed in such methods are relatively limited. For example, in the anionic polymerization of styrene and butadiene to form a SBS type block copolymer, each polymer chain requires a stoichiometric amount of initiator and the resulting polymers have extremely narrow molecular weight distribution, Mw/Mn, preferably from 1.0 to 1.3. That is, the polymer block lengths are substantially identical. Additionally, anionic and free-radical processes are relatively slow, resulting in poor process economics, and not readily adapted to polymerization of α-olefins.
Previous researchers have stated that certain homogeneous coordination polymerization catalysts can be used to prepare polymers having a substantially “block-like” structure by suppressing chain-transfer during the polymerization, for example, by conducting the polymerization process in the absence of a chain transfer agent and at a sufficiently low temperature such that chain transfer by β-hydride elimination or other chain transfer processes is essentially eliminated. Under such conditions, the sequential addition of different monomers coupled with high conversion was said to result in formation of polymers having sequences or segments of different monomer content. Several examples of such catalyst compositions and processes are reviewed by Coates, Hustad, and Reinartz in Angew. Chem., Int. Ed., 41, 2236-2257 (2002) as well as US-A-2003/0114623.
Disadvantageously, such processes require sequential monomer addition and result in the production of only one polymer chain per active catalyst center, which limits catalyst productivity. In addition, the requirement of relatively low process temperatures but high conversion increases the process operating costs, making such processes unsuited for commercial implementation. Moreover, the catalyst cannot be optimized for formation of each respective polymer type, and therefore the entire process results in production of polymer blocks or segments of less than maximal efficiency and/or quality. For example, formation of a certain quantity of prematurely terminated polymer is generally unavoidable, resulting in the forming of blends having inferior polymer properties.
The use of certain metal alkyl compounds and other compounds, such as hydrogen, as chain transfer agents to interrupt chain growth in olefin polymerizations is well known in the art. In addition, it is known to employ such compounds, especially aluminum alkyl compounds, as scavengers or as cocatalysts in olefin polymerizations. In Macromolecules, 33, 9192-9199 (2000) the use of certain aluminum trialkyl compounds as chain transfer agents in combination with certain paired zirconocene catalyst compositions resulted in polypropylene mixtures containing small quantities of polymer fractions containing both isotactic and atactic chain segments. In Liu and Rytter, Macromolecular Rapid Comm., 22, 952-956 (2001) and Bruaseth and Rytter, Macromolecules, 36, 3026-3034 (2003) mixtures of ethylene and 1-hexene were polymerized by a similar catalyst composition containing trimethylaluminum chain transfer agent. In the latter reference, the authors summarized the prior art studies in the following manner (some citations omitted):
“Mixing of two metallocenes with known polymerization behavior can be used to control polymer microstructure. Several studies have been performed of ethene polymerization by mixing two metallocenes. Common observations were that, by combining catalysts which separately give polyethene with different Mw, polyethene with broader and in some cases bimodal MWD can be obtained. [S]oares and Kim (J. Polym. Sci., Part A: Polym. Chem., 38, 1408-1432 (2000)) developed a criterion in order to test the MWD bimodality of polymers made by dual single-site catalysts, as exemplified by ethene/1-hexene copolymerization of the mixtures Et(Ind)2ZrCl2/Cp2HfCl2 and Et(Ind)2ZrCl2/CGC (constrained geometry catalyst) supported on silica. Heiland and Kaminsky (Makromol. Chem., 193, 601-610 (1992)) studied a mixture of Et-(Ind)2ZrCl2 and the hafnium analogue in copolymerization of ethene and 1-butene.
These studies do not contain any indication of interaction between the two different sites, for example, by readsorption of a terminated chain at the alternative site. Such reports have been issued, however, for polymerization of propene. Chien et al. (J. Polym. Sci., Part A: Polym. Chem., 37, 2439-2445 (1999), Makromol., 30, 3447-3458 (1997)) studied propene polymerization by homogeneous binary zirconocene catalysts. A blend of isotactic polypropylene (i-PP), atactic polypropylene (a-PP), and a stereoblock fraction (i-PP-b-a-PP) was obtained with a binary system comprising an isospecific and an aspecific precursor with a borate and TIBA as cocatalyst. By using a binary mixture of isospecific and syndiospecific zirconocenes, a blend of isotactic polypropylene (i-PP), syndiotactic polypropylene (s-PP), and a stereoblock fraction (i-PP-b-s-PP) was obtained. The mechanism for formation of the stereoblock fraction was proposed to involve the exchange of propagating chains between the two different catalytic sites. Przybyla and Fink (Acta Polym., 50, 77-83 (1999)) used two different types of metallocenes (isospecific and syndiospecific) supported on the same silica for propene polymerization. They reported that, with a certain type of silica support, chain transfer between the active species in the catalyst system occurred, and stereoblock PP was obtained. Lieber and Brintzinger (Macromol. 3, 9192-9199 (2000)) have proposed a more detailed explanation of how the transfer of a growing polymer chain from one type of metallocene to another occurs. They studied propene polymerization by catalyst mixtures of two different ansa-zirconocenes. The different catalysts were first studied individually with regard to their tendency toward alkyl-polymeryl exchange with the alkylaluminum activator and then pairwise with respect to their capability to produce polymers with a stereoblock structure. They reported that formation of stereoblock polymers by a mixture of zirconocene catalysts with different stereoselectivities is contingent upon an efficient polymeryl exchange between the Zr catalyst centers and the Al centers of the cocatalyst.”
Brusath and Rytter then disclosed their own observations using paired zirconocene catalysts to polymerize mixtures of ethylene/1-hexene and reported the effects of the influence of the dual site catalyst on polymerization activity, incorporation of comonomer, and polymer microstructure using methylalumoxane cocatalyst.
Analysis of the foregoing results indicate that Rytter and coworkers likely failed to utilize combinations of catalyst, cocatalyst, and third components that were capable of readsorption of the polymer chain from the chain transfer agent onto both of the active catalytic sites, that is they failed to obtain two-way readsorption. While indicating that chain termination due to the presence of trimethylaluminum likely occurred with respect to polymer formed from the catalyst incorporating minimal comonomer, and thereafter that polymeryl exchange with the more open catalytic site followed by continued polymerization likely occurred, evidence of the reverse flow of polymer ligands appeared to be lacking in the reference. In fact, in a later communication, Rytter, et. al., Polymer, 45, 7853-7861 (2004), it was reported that no chain transfer between the catalyst sites actually took place in the earlier experiments. Similar polymerizations were reported in WO98/34970.
In U.S. Pat. Nos. 6,380,341 and 6,169,151, use of a “fluxional” metallocene catalyst, that is a metallocene capable of relatively facile conversion between two stereoisomeric forms having differing polymerization characteristics such as differing reactivity ratios was said to result in production of olefin copolymers having a “blocky” structure. Disadvantageously, the respective stereoisomers of such metallocenes generally fail to possess significant difference in polymer formation properties and are incapable of forming both highly crystalline and amorphous block copolymer segments, for example, from a given monomer mixture under fixed reaction conditions. Moreover, because the relative ratio of the two “fluxional” forms of the catalyst cannot be varied, there is no ability, using “fluxional” catalysts, to vary polymer block composition or to vary the ratio of the respective blocks.
In JACS, 2004, 126, 10701-10712, Gibson, et al discuss the effects of “catalyzed living polymerization” on molecular weight distribution. The authors define catalyzed living polymerization in this manner:                “ . . . if chain transfer to aluminum constitutes the sole transfer mechanism and the exchange of the growing polymer chain between the transition metal and the aluminum centers is very fast and reversible, the polymer chains will appear to be growing on the aluminum centers. This can then reasonably be described as a catalyzed chain growth reaction on aluminum . . . . An attractive manifestation of this type of chain growth reaction is a Poisson distribution of product molecular weights, as opposed to the Schulz-Flory distribution that arises when β-H transfer accompanies propagation.”        
The authors reported the results for the catalyzed living homopolymerization of ethylene using an iron containing catalyst in combination with ZnEt2, ZnMe2, or Zn(i-Pr)2. Homoleptic alkyls of aluminum, boron, tin, lithium, magnesium and lead did not induce catalyzed chain growth. Using GaMe3 as cocatalyst resulted in production of a polymer having a narrow molecular weight distribution. However, after analysis of time-dependent product distribution, the authors concluded this reaction was, “not a simple catalyzed chain growth reaction.” Similar processes employing similar catalysts have been described in U.S. Pat. Nos. 5,210,338, 5,276,220, and 6,444,867.
Earlier workers had made claims to forming block copolymers using a single Ziegler-Natta type catalyst in multiple reactors arranged in series. Examples of such teachings include U.S. Pat. Nos. 3,970,719 and 4,039,632. It is now known that no substantial block copolymer formation takes place under these reaction conditions.
It is known in the art that the presence of long chain branching (LCB) may improve certain polymer characteristics, especially processability and melt strength. The presence of LCB in a polymer is characterized by the occurrence of polymer moieties of a length greater than that of any C3-8 olefin comonomer remnant attached to the main, backbone polymer chain. In prior art techniques, long chain branching may be generated in a polymer by incorporation of a vinyl-terminated macromer (either deliberately added or formed in situ during a polymerization such as through β-hydride elimination) either by action of the polymerization catalyst itself or by the use of a linking agent. These methods generally suffer from incomplete incorporation of the vinyl-terminated macromer or linking moiety into the polymer, and/or a lack of control over the extent of LCB for given process conditions.
Accordingly, there remains a need in the art for a polymerization process that is capable of preparing copolymers having unique properties in a high yield process adapted for commercial utilization. Moreover, it would be desirable if there were provided an improved process for preparing polymers, including copolymers of two or more comonomers such as ethylene and one or more comonomers, by the use of a polymerizable shuttling agent (PSA) to introduce branching, including long chain branching, in the resulting copolymers, especially pseudo-block copolymers. It would also be desirable to provide a method for generating controlled amounts of long chain branching in olefin polymers, especially pseudo-block copolymers, that does not require the in situ formation of vinyl functionalized macromolecules. In addition it would be desirable to provide such an improved process for preparing the foregoing branched copolymer products in a continuous polymerization process.